Hybrid Video Compression Standard [1st ed. 2020] 978-981-15-0244-6, 978-981-15-0245-3

Table of contents :
Front Matter ....Pages i-xiii
Introduction to Video Compression (Dhaval R. Bhojani, Vedvyas J. Dwivedi, Rohit M. Thanki)....Pages 1-14
Technical Background (Dhaval R. Bhojani, Vedvyas J. Dwivedi, Rohit M. Thanki)....Pages 15-27
Standard Video Codec (Dhaval R. Bhojani, Vedvyas J. Dwivedi, Rohit M. Thanki)....Pages 29-42
Hybrid Video Codec (Dhaval R. Bhojani, Vedvyas J. Dwivedi, Rohit M. Thanki)....Pages 43-56
Comparative Comparison of Standard and Hybrid Video Codec (Dhaval R. Bhojani, Vedvyas J. Dwivedi, Rohit M. Thanki)....Pages 57-58

Citation preview

SPRINGER BRIEFS IN APPLIED SCIENCES AND TECHNOLOGY  COMPUTATIONAL INTELLIGENCE

Dhaval R. Bhojani Vedvyas J. Dwivedi Rohit M. Thanki

Hybrid Video Compression Standard

123

SpringerBriefs in Applied Sciences and Technology Computational Intelligence

Series Editor Janusz Kacprzyk, Polish Academy of Sciences, Systems Research Institute, Warsaw, Poland

SpringerBriefs in Computational Intelligence are a series of slim high-quality publications encompassing the entire spectrum of Computational Intelligence. Featuring compact volumes of 50 to 125 pages (approximately 20,000-45,000 words), Briefs are shorter than a conventional book but longer than a journal article. Thus Briefs serve as timely, concise tools for students, researchers, and professionals.

More information about this subseries at http://www.springer.com/series/10618

Dhaval R. Bhojani Vedvyas J. Dwivedi Rohit M. Thanki •

•

Hybrid Video Compression Standard

123

Dhaval R. Bhojani Electronics and Communication Government Engineering College Rajkot, Gujarat, India

Vedvyas J. Dwivedi C. U. Shah University Wadhwan City, Gujarat, India

Rohit M. Thanki C. U. Shah University Wadhwan City, Gujarat, India

ISSN 2191-530X ISSN 2191-5318 (electronic) SpringerBriefs in Applied Sciences and Technology ISSN 2625-3704 ISSN 2625-3712 (electronic) SpringerBriefs in Computational Intelligence ISBN 978-981-15-0244-6 ISBN 978-981-15-0245-3 (eBook) https://doi.org/10.1007/978-981-15-0245-3 © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

The usage of social media and sharing of multimedia content over the Internet are rapidly increasing in recent time. Various facilities such as sharing of videos, downloading and viewing them, and videoconferencing based on the Internet are routine in the modern lifestyle. These all applications are consuming high channel bandwidth and capacity, but most of the channels have limited bandwidth. Therefore, the biggest problem for transmission of a high amount of video content is that it cannot be transmitted over a small bandwidth channel. So, to overcome this problem, compression of video is necessary before transmitting it. Video is a sequence of image frames with a large amount of redundancy between two frames. This redundancy may be subjective or statistical. The main aim of video compression techniques is to reduce this redundancy using different types of mathematical and information technology-related techniques. The output of these techniques is compressed video content with reduced redundancy than in original video content. Based on the applications, video compression techniques are divided into two types: “lossy” and “lossless.” The main aim of lossless video compression technique is reducing the data in video content for sharing and storing without losing information in video content. While lossy video compression techniques do the same but lose information in video content, and the same concept is envisioned by various video standards such as Moving Picture Experts Group (MPEG) with different versions such as MPEG-1, MPEG-2, and MPEG-4. These standards are widely used for the application of video content transmission over a communication channel which has less bandwidth and limited storage capacity. The main challenges faced by these standards are no good trade-off between quality of video content, algorithm complexity, and achieved compression ratio. The latest MPEG-4 standard has good algorithm complexity and objective-based algorithm, while MPEG-2 has less algorithm complexity and subjective-based algorithm. This book aims to provide basic video compression standards, particularly MPEG with its different versions. The detailed working of standard video codec with the help of figures is given in this book, so that readers can easily get the concept of this basic video standard. After that, the hybrid video codec is explained with details to achieve improved performance compared to standard video codec v

vi

Preface

without including the computation complexity of the algorithm. Finally, the comparison of the obtained results for these standards is discussed.

Overview of the Book In Chap. 1, basic information and properties of various video compression standards are briefly discussed. The rest of the book covers various video compression standards in Chaps. 2–4 along with technical background for video compression, working and experimental results of MPEG-2 and modified MPEG-2 standards. In Chap. 2, various terminologies such as color conversion, estimation of motion, transform coding, quantization, zigzag scanning, and entropy coding that are used in video compression are discussed. The working of the standard video codec is covered in Chap. 3. In this chapter, the basic concept of video encoder and decoder along with experimental results is demonstrated with the help of real-time video signal. The working of the hybrid video codec is covered in Chap. 4. Finally, Chap. 5 gives a comparison of the obtained results of these standards.

Features of the Book • • • •

Basic information on video compression standards Detailed working of standard video codec for compression of the video signal Extensive discussion on a modified version of standard video codec Inclusion of results of video compression for real-time video signals

Acknowledgements My task has been easier and the final version of the book is considerably better because of the help we have received. Acknowledging that help is itself a pleasure. I would extend many thanks to all persons who helped achieving the final version of this book. The authors are indebted to numerous colleagues for valuable suggestions during the entire period of the manuscript preparation. I would also like to thank the staff at Springer, in particular Aninda Bose, senior publishing editor/CS Springer, for their helpful guidance and encouragement during the creation of this book. Rajkot, Gujarat, India Wadhwan City, Gujarat, India Rajkot, Gujarat, India

Dr. Dhaval R. Bhojani Dr. Vedvyas J. Dwivedi Dr. Rohit M. Thanki

Contents

1 Introduction to Video Compression . . . . . . . . . . . . . . . 1.1 Basic of Video Compression . . . . . . . . . . . . . . . . . . 1.2 Types and Need for Video Compression Techniques 1.3 Problems in Video Compression . . . . . . . . . . . . . . . 1.4 Various Video Compression Standards . . . . . . . . . . 1.4.1 Motion JPEG . . . . . . . . . . . . . . . . . . . . . . . 1.4.2 H.26x . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4.3 MPEG-x . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.5 Applications of Video Compression Standards . . . . . 1.6 Video Compression Model . . . . . . . . . . . . . . . . . . . 1.7 Literature Survey on Video Compression Standards . 1.8 Points Cover in the Book . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

1 2 4 5 6 6 6 7 8 8 10 12 12

2 Technical Background . . . . . . . . . . . . . . 2.1 Color Space Conversion . . . . . . . . . . 2.2 Re-sampling of Chrominance . . . . . . 2.3 Motion Estimation and Compensation 2.4 Transformation Coding . . . . . . . . . . . 2.5 Quantization and Zigzag Scanning . . 2.6 Entropy Coding . . . . . . . . . . . . . . . . 2.7 Video Quality Matrices . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

15 15 16 17 19 23 24 25 27

3 Standard Video Codec . . . . . . . . . . . . . . . . . . . . . 3.1 Video Compressor . . . . . . . . . . . . . . . . . . . . . 3.2 Video Decompressor . . . . . . . . . . . . . . . . . . . 3.3 Experimental Results of Standard Video Codec 3.3.1 Results for Foreman Video Signal . . . . 3.3.2 Results for Vipman Video Signal . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

. . . . . .

29 29 30 31 32 32

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

. . . . . . . . .

vii

viii

Contents

3.4 Summary of Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 4 Hybrid Video Codec . . . . . . . . . . . . . . . . . . . . . . 4.1 Hybrid Video Compressor . . . . . . . . . . . . . . 4.2 Hybrid Video Decompressor . . . . . . . . . . . . . 4.3 Experimental Results of Hybrid Video Codec 4.3.1 Results for Foreman Video Signal . . . 4.3.2 Results for Vipman Video Signal . . . . 4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

43 43 45 46 46 47 51 56

5 Comparative Comparison of Standard and Hybrid Video Codec . . . . 57

About the Authors

Dr. Dhaval R. Bhojani is an Assistant Professor, Department of E.C. Engineering, Government Engineering College, Rajkot. He earned hist PhD in video compression from JJTU, Rajasthan in 2013. His areas or research are Image & video processing, industrial automation and embedded systems. He has published more than 20 research papers in refereed and indexed journals, and has participated in conferences at the international and national level. He has member of ISTE and Institute of Engineers, New Delhi, India. Dr. Vedvyas J. Dwivedi is a Professor (Department of E.C. Engineering, Faculty of Technology and Engineering) and Vice-Provost of C. U. Shah University. He has guided 7 Ph.D. theses, 43 M. Tech. dissertations, examined 8 Ph.D. theses from Indian Universities, 4 Ph.D. Theses from the United States based Universities, published 137 research and review articles, delivered 38 expert/resource talks, authored/co-authored 10 books, completed 12 research/consultancy projects, published 8 patents, chaired 16 sessions in national/international conferences of IEEE, IETE, IE (I), ISTE and HEIs. His expertise and interest areas are wireless-satellite-mobile-optical-RF-Microwave systems, sensor-energy-signal technology. His videos on YouTube and www.cushahuniversity.ac.in are available. He has earned B.E (Electronics Engineering), M.E. (Electronics and Communication Engineering), and Ph.D. (Electronics and Communication Engineering). Dr. Rohit M. Thanki received his Ph.D. in electronics and communication engineering from C. U. Shah University, M.E. in communication engineering from G. H. Patel College of Engineering and Technology and B.E. in electronics and communication engineering from Atmiya Institute of Technology and Science, India. He has more than 3 years of experience in academic and research. He has published 10 books with Springer and 1 book with CRC press. He has published 13 book chapters in edited books which are published by Elsevier, Springer, CRC press, and IGI Global. He has also published 20 research articles, out of these, 6 articles in SCI-indexed journals and 14 articles in Scopus indexed journals. He is a ix

x

About the Authors

reviewer of renowned journals such as IEEE access, IEEE Consumer Electronics Magazine, IET Image Processing, IET Biometrics, Soft Computing, Imaging Science Journal, Signal Processing: Image Communication, and Computers & Electrical Engineering. His current research interests include Image Processing, Multimedia Security, Digital Watermarking, Artificial Intelligence, Medical Image Analysis, Biometrics, and Compressive Sensing.

List of Figures

Fig. Fig. Fig. Fig.

1.1 1.2 1.3 1.4

Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig.

1.5 1.6 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10

Fig. Fig. Fig. Fig. Fig. Fig.

2.11 3.1 3.2 3.3 3.4 3.5

Fig. 3.6 Fig. 3.7

Data compression model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example of spatial redundancy . . . . . . . . . . . . . . . . . . . . . . . . . . Example of temporal redundancy . . . . . . . . . . . . . . . . . . . . . . . . Trade-off between quality of video and compression rate (CR). a Original frame (24.9 kB). b Frame with a low compression rate (3.65 kB). c Compressed frame with very CR (2.19 kB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . History of video compression standards. . . . . . . . . . . . . . . . . . . . Video compression model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example of color space conversion . . . . . . . . . . . . . . . . . . . . . . . Example of re-sampling of chrominance components . . . . . . . . . Prediction of frames in the Group of Pictures (GOP) . . . . . . . . . . Block matching process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Estimation of error in frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . a Basis of DCT. b Frequency distribution in DCT block . . . . . . Example of discrete wavelet transform (DWT) . . . . . . . . . . . . . . a Original frame. b Truncated frame . . . . . . . . . . . . . . . . . . . . . . Standard quantization matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . Example of quantization process. a Original coefficients. b Process coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zigzag scanning process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Block diagram of video compressor . . . . . . . . . . . . . . . . . . . . . . Block diagram of video decompressor . . . . . . . . . . . . . . . . . . . . . Original foreman video frames . . . . . . . . . . . . . . . . . . . . . . . . . . Compressed foreman video frames for standard video codec . . . . Quality metrices for each frames of foreman video signal for standard video codec. a PSNR values, b MSE values, c SSIM values, d MSAD values, e VQM values, f blocking beta values, and g blurring beta values . . . . . . . . . . . . . . . . . . . . . . . . Original vipman video frames . . . . . . . . . . . . . . . . . . . . . . . . . . . Compressed vipman video frames for standard video codec . . . .

1 3 3

4 6 9 16 17 18 19 20 21 22 23 23 24 25 30 31 33 34

35 38 39 xi

xii

List of Figures

Fig. 3.8

Fig. Fig. Fig. Fig. Fig.

4.1 4.2 4.3 4.4 4.5

Fig. 4.6 Fig. 4.7 Fig. 4.8

Fig. 5.1

Quality metrics for each frames of vipman video signal for standard video codec. a PSNR values, b MSE values, c SSIM values, d MSAD values, e VQM values, f blocking beta values, and g blurring beta values . . . . . . . . . . . . . . . . . . . . . . . . Block diagram of hybrid video compressor . . . . . . . . . . . . . . . . . Block diagram of hybrid video decompressor . . . . . . . . . . . . . . . Original foreman video frames . . . . . . . . . . . . . . . . . . . . . . . . . . Compressed foreman video frames for hybrid video codec . . . . . Quality metrics for each frames of foreman video signal for hybrid video codec a PSNR values, b MSE values, c SSIM values, d MSAD values, e VQM values, f blocking beta values, and g blurring beta values . . . . . . . . . . . . . . . . . . . . . . . . Original vipman video frames . . . . . . . . . . . . . . . . . . . . . . . . . . . Compressed vipman video frames for hybrid video codec . . . . . . Quality metrics for each frames of vipman video signal for hybrid video codec a PSNR values, b MSE values, c SSIM values, d MSAD values, e VQM values, f blocking beta values, and g blurring beta values . . . . . . . . . . . . . . . . . . . . . . . . Comparative comparison of video compression standards a for foreman video signal and b for vipman video signal . . . . . . . . . .

40 44 45 47 48

49 52 53

54 58

List of Tables

Table Table Table Table

1.1 1.2 1.3 3.1

Table 3.2 Table 4.1 Table 4.2

Information on formats of various video signals . . . . . . . . . Applications of video signal with its data rates . . . . . . . . . . Applications of video compression standards . . . . . . . . . . . Quality matrices for compressed foreman video signal for standard video codec . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality matrices for compressed vipman video signal for standard video codec . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality matrices for compressed foreman video signal for hybrid video codec . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quality matrices for compressed vipman video signal for hybrid video codec . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.... .... ....

2 5 9

....

34

....

39

....

48

....

53

xiii

Chapter 1

Introduction to Video Compression

Keywords Compression · MPEG · Video · PSNR After the last of the 1940s, data compression has been important terminology and widely used in the field of information sharing and distribution. The technique related to data compression aims to reduce redundancy presented in the data [1]. The compression technique has two types of blocks such as encoder and decoder which performs encoding and decoding of data in such a way the reduce redundancy in the data. The basic model compression technique is shown in Fig. 1.1 which contains encoder or compressor, communication channel, and decoder or decompressor. The work of encoder or compressor is encoded and reduced the data. This compressed data may be transmitted through the communication channel and given as input to decoder or decompressor. The work of decoder or decompressor is to get original data from the compressed data. Here, the ratio between the data rate of the encoder and the data rate of the channel is referred to as the compression ratio. When working of the encoder is more complex than working of decoder, then the type of model is referred to as asymmetrical [2]. The motion picture expert group (MPEG) is working on this type of model. The data compression has a property like that it transforms data into a string of symbols which would require lesser bits compared to the representation of original data. The types of data such as text, audio, image, and video are transmitted on any communication channel. Out of these data, the videos are used more and more in today life and people are sharing it using various applications and devices.

Fig. 1.1 Data compression model © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 D. R. Bhojani et al., Hybrid Video Compression Standard, SpringerBriefs in Computational Intelligence, https://doi.org/10.1007/978-981-15-0245-3_1

1

2

1 Introduction to Video Compression

Table 1.1 Information on formats of various video signals S.No.

Video format

Resolution

1

Quarter Common Intermediate Format (QCIF)

176 × 144

Require storage capacity (Mbps) 17.4

2

Common Intermediate Format (CIF)

352 × 288

69.61

3

Phase Alternating Line (PAL)

720 × 576

284.77

4

National Television System Committee (NTSC)

720 × 480

237.3

5

High Definition Television (HDTV)

1280 × 720

632.81

The digital videos can be considered as sequences of the image frame, and each frame has some amount of information in term of pixels. The value of these pixels defined the resolution of the video frame. Table 1.1 shows the resolution of various video formats with its required storage requirement [3–5]. These storage capacities require for storage of color video stream which has 30 frames per second capture rate. As per referring to Table 1.1, it is indicated a small amount of video has required a large amount of data storage capacity. So, the large memory is required in any system which is built based on the video. Therefore, digital video compression algorithm represents video in a compressed manner with a smaller size compared to the original version and retains the quality of the original video in terms of perceptibility. It is working of the principle of perceptibility capacity of the human. The human eye cannot be identified as changes in chrominance components of color video. This property is widely used in many video compression standards which remove redundancy in video content to get its compressed version. The basic of video compression is discussed in the next subsection.

1.1 Basic of Video Compression Video can be represented as a three-dimensional array with various digits, where the first two dimensions represent horizontal and vertical while the third dimension is the time coordinate. The video frame comprises of color pixels that correspond to a single time moment of a video. This frame is reference as video frame [6]. The video data has different types of redundancy in various domains such as spatial and temporal. Also, some other types of redundancies such as perceptual and statistical are associated with video data [7]. The similarities within a frame refer as spatial redundancy while similarities between two consecutive frames refer as temporal redundancy. The human eye cannot distinguish small changes in color components while it is clearly identifying

1.1 Basic of Video Compression

3

Fig. 1.2 Example of spatial redundancy

changes in brightness of similar color which has more spatial redundancy. The intraframe coding is used reducing this type redundancy within the frame. The example of spatial redundancy is shown in Fig. 1.2. This type of redundancy can be reduced by various types of coding such as run-length, Huffman. For reducing temporal redundancy, changes from one video frame to the next video frame identifies and using temporal compression encoded these changes. Here, temporal compression divides the whole video frame into nonoverlapping blocks and compression performs on them. These blocks refer to macroblocks. These blocks of one frame are then compared to the blocks of the next frame and encoder sends only differences between these blocks to reduce the redundancy. The example of temporal redundancy shows in Fig. 1.3 which shows a few frames of video having very minor changes from one frame to another frame. So, most of the temporal redundancy can be removed before transmission of it, which requires a small transmission data rate for the same video [4]. The perceptual redundancy refers that the details in video data which cannot perceive by the human eye. The information which can’t perceive by human eye can be removed by the compression technique without affecting the quality of video data. The statistical redundancy is associated with different technique terminologies such as transform coefficients, vectors related to motion, codes. Therefore, a suitable chosen of this technique will get proper compression in video data.

Fig. 1.3 Example of temporal redundancy

4

1 Introduction to Video Compression

1.2 Types and Need for Video Compression Techniques The video compression techniques divide into two types such as lossy compression and lossless compression. In lossless compression, when the compressed video is decompressed, then resultant video has perfect matching with original video in the spatial domain. This type of compression is mainly removing spatial redundancy using intra-coding method. This type of techniques is rarely used due to less compression ratio [5, 6]. In lossy compression, when the compressed video is decompressed then resultant video has not given preface matching with original video in the spatial domain. This type of compression gives high compression ratio which predicts the changes in the frame using motion estimation and finds residue between two consecutive frames to reduced temporal redundancy in the frame. Based on the coding method, compression divides into two types such as intracoded and inter-coded. The intra-coding method works on three types of data such as x-direction and y-direction and sample value lie on location (x, y). The standard TV video signal contains high-frequency components in it due to a large number of details. Also, some information in small portion has similar pixel value which increases low-frequency components in it. The brightness of the video depends on zero frequency components. Therefore, suitable coding methods are required which work effectively in all types of frequencies in the video signal. The various signal transforms such as wavelet, discrete cosine transform (DCT) are used with the intra-coding method for suitable video compression. The inter-coding method works on similarities between consecutive frames of the video signal. This method sends only information about these similarities and reconstructed compression video using various differential coding. The video compression technique has a trade-off between quality of the video, storage capacity, and cost of implementation. If the video compressed using the lossy technique, visible changes appear in the resultant video. Figure 1.4 shows the frame of the original video and its various compression version with different size of compression. From Fig. 1.4, it is seen that the compression increases storage capacity but decreases the quality of the video frame. After some rate of compression, the

Fig. 1.4 Trade-off between quality of video and compression rate (CR). a Original frame (24.9 kB). b Frame with a low compression rate (3.65 kB). c Compressed frame with very CR (2.19 kB)

1.2 Types and Need for Video Compression Techniques

5

Table 1.2 Applications of video signal with its data rates S. No.

Applications

Video resolution and frame per second (fps)

Uncompressed data rate

Compressed data rate

1

Videoconferencing

352 × 240 and 15

30.4 Mbps

64–768 kbps

2

CD-ROM digital video

352 × 240 and 30

60.8 Mbps

1.5–4 Mbps

3

TV broadcasting

720 × 480 and 30

248.8 Mbps

3–8 Mbps

4

HDTV

1280 × 720 and 60

1.33 Gbps

20 Mbps

quality of the video frame reduces such a way that the information of the frame cannot properly visible by the human eye. Table 1.2 shows the applications of the video signal in real life with the required data rate for transmission of video signals and its compressed format. The table shows that due to compression, less amount of channel capacity required for transmission of this type of video signal and less memory for storage of it. This is one of the motivations behind development for various types of video compression standards.

1.3 Problems in Video Compression The various types of problems are associated with video compression standard which is discussed below: • The quality of the video signal is very sensitive, and compression method affects this quality parameter. • The dropping of the video frame may happen during transmission of it. If the compression is not applied on individual framer, then some of the data may be lost after compression of it and proper reconstruction cannot be performed at the receiver side. • The complexity of existing video compression standards is high. • The video signal requires more bandwidth and so that it cannot be transmitted over a low bandwidth channel. Therefore, the compression of the video signal is required. To overcome these problems in video compression, the hybrid video compression standard is discussed in this book which decreases the requirement of channel capacity and storage capacity while keeping video quality is higher.

6

1 Introduction to Video Compression

1.4 Various Video Compression Standards The various agencies worldwide are developed and defined various standards for video coding and compression. These standards are developed by two agencies which names are video coding expert group (VCEG) of telecommunication standardization sector of international telecommunication union (ITU-T) and moving picture expert group (MPEG) of the international organization for standardization (ISO/IEC). The standards developed by VCEG are called as H.26x and developed by MPEG are called as MPEG-x [7–11]. The history of these video compression standards is shown in Fig. 1.5. The information of these standards is given as below.

1.4.1 Motion JPEG The video signal is nothing but a sequence of video frame which is treated as an image. Therefore, in this standard, taken the advantage of JPEG compression standards such as JPEG and JPEG 2000 and applied on sequences of images to achieve suitable quality and compression ratio. These standards refer as MJPEG and MJPEG 2000. The main disadvantage of these standards is that there are image compression standards and only applicable for sequences of images which cannot be treated as a video compression method [12].

1.4.2 H.26x The details of various H.26x standards are given as below:

Fig. 1.5 History of video compression standards

1.4 Various Video Compression Standards

7

• H.261: This standard was developed by the ITU-T group around 1990. This standard was designed for the transmission of video signals with a data rate of 64 kbps and its multiplying values. The frame of this standard can be divided into two types such as I (intra-coded) frame and P (predicted) frame. In I frame, coding of the frame is done without information of the previous frame where coding of the frame is done using the information of the previous frame in P frame. It is used motioncompensation temporal prediction to find similarities between two frames. This standard widely used in applications like video calling and videoconferencing. • H.263: This standard was developed by the ITU-T group around 1995. This standard was designed for lower data rates. In this standard, the video frame was divided into number of macroblocks which have 16 × 16 blocks of the luminance component and 8 × 8 blocks of chrominance blocks. These blocks were encoded as intra- or inter-coding. In this standard, spatial redundancy is explored by DCT coding and temporal redundancy is explored by motion compensation. This standard encoded the signal with an accuracy of half-pixel and bidirectional coding. This standard provides a low bit rate which can be provided by standards like MPEG-1 and MPEG-2. • H.264: This standard was developed with ITU-T together with the ISO/IEC group. This standard based on block-oriented motion compensation with providing a lower bit rate than existing standards such as H.263, MPEG-2, and MPEG-4. This standard provides flexibility in various applications related to networking and systems such as video broadcasting, storage of video in DVD, multimedia transmission and telephonic system which low and high bit rates are required. It widely used as a lossy compression standard but also provides lossless compression standard features. This standard provides 1.5 Mbps for transmission of digital satellite TV signal while MPEG-2 requires around 3.5 Mbps for the same signal. • H.265: This standard refers to high-efficiency video coding standard and widely used for compression of the video signal. The aim behind designing this standard is that it provides higher coding capacity at a lower bit rate. It provides support to high-resolution video signal with maximum resolution up to 8192 × 4329. It provides a double compression ratio compared to other standards with same bit rate and video quality.

1.4.3 MPEG-x The details of various MPEG-x standards are given as below: • MPEG-1: This standard was developed by the MPEG group around 1991. This was the first standards developed by MPEG and accepted worldwide. It compressed video quality with a maximum allowed bit rate of 1.5 Mbits/s. The standard operates on data rate between 1 and 2 Mbit/s. This standard used motion estimation and compensation at 16 × 16 macroblocks with full-motion search as a coding

8

1 Introduction to Video Compression

method which was used by H.261 standard. Additionally, this standard use blockwise discrete cosine transform, quantization, and entropy coding. In addition to H.261, this standard added bidirectionally predicated frame and half-pixel motion search. Due to additional features, this standard improves the data rate of 1 Mbits/s compared to H.261 standard. • MPEG-2: This standard was developed by the MPEG group around 1993. MPEG2 standards mainly designed for achieved high compression ratio in the TV broadcasting signal. The targeted compression rate for this standard is 4–30 Mbps with providing high quality to the video signal. The additional feature in this standard is that it used the coding of interlace scanned picture which was for stability in the video signal and makes effectively transmission of it in various channels. The standard also introduced new feature such as video codecs’ concept which provides scalability in video signals and became a standard feature for another video compression standard. • MPEG-4: This standard was developed by the MPEG group around 1993. MPEG4 adopted new coding features such as motion estimation and compensation with variable block size, variable-length coding-based entropy coding, error removal capability for transmission of the video signal in compressed format. This standard supports the transmission of the video signal in lower bandwidth channel with high quality. The compression ratio of 100:1 is common for this standard. The standard mainly uses in various applications such as mobile communication where low bandwidth channels are used.

1.5 Applications of Video Compression Standards The main application of video compression standards is to compressed video signals which are transmitted for TV broadcasting and required low storage or memory on the disk. The other possible applications of video compression are used in the broadcasting of UHF and VHF video signals, satellite TV broadcasting, manufacturing of VCD and DVD, videoconferencing, and transmission of high definition TV (HDTV) signals. Table 1.3 shows applications of various video compression standards with its required data rate.

1.6 Video Compression Model The generalized block diagram for a video compression model is given in Fig. 1.6. This model consists of two types of blocks such as video encoder and video decoder. Here, a video frame encoded by encoder which converts the frame into a set of bits. After transmission of bits through the channel, the encoded bits are given to input to a decoder where the reconstructed frame is generated. If the channel is not error-free,

1.6 Video Compression Model

9

Table 1.3 Applications of video compression standards S. No.

Standards

Applications

Data rate

1

MJPEG and MJPEG 2000

Sequences of still images

Not specified

2

H.261

Videoconferencing and telephony

64 kbps

3

MPEG-1

Storage of video on DVD/CD-ROM

1.5 Mbps

4

MPEG-2

TV broadcasting

>2 Mbps

5

H. 263

Videoconferencing